APPLICATION OF TABU SEARCH AND SIMULATED ANNEALING BASED ALGORITHMS FOR SOLUTION OF FMS LOADING PROBLEMS

Abstract

This work aims at solving machine loading problem of FMS (Flexible Manufacturing System) with the bicriterion objective of minimizing system unbalance and maximizing throughput in presence of technological constraints such as available machining time and tool slots. Tabu search and Simulated annealing based algorithms are employed to search the solution of above problems. These algorithms work in two phases. In the first phase initial seed sequence generation is carried out from the fixed job sequencing rules. A methodology based on the utilization of maximum remaining available time on each machine while keeping in mind the technological constraints for achieving the minimum system unbalance and maximum throughput has been adopted to rejected jobs. In the second phase Tabu search and simulated annealing based algorithms have been pursued to improve the results of first stage. Proposed methodologies have been tested on Ten problems and results obtained are compared with some existing heuristics, which have also been shown in different tables. Results are better for majority of test problems in underlying objectives i.e. minimizing of system unbalance and maximizing of throughput The programming of these algorithms is written in C++ and runs in PC/386 compatible under DOS and UNIX.